Carrie Hersh, DO, MSc, FAAN on LinkedIn: February 2024 (2024)

This article by Sims et al. (2023) provides insight into the myriad of various ways that students, particularly those who are first in their immediate families aspiring to become medical professionals, navigate the journey to medical school. In addition, the authors highlight the ways in which complex experiences, contexts, and considerations are embedded within and influence their jouneys. Link: https://lnkd.in/g2ekwGuf

🔬Exciting News in Medical AI! 🔬As someone involved in data annotation for cutting-edge projects, I've been captivated by the latest updates from Google. As I delved into their updates on medical summaries, I couldn't help but be amazed by the strides they've made.Their health research teams have utilized the power of PaLM 2 to create Med-PaLM 2, a remarkable tool fine-tuned for medical knowledge. This advancement enables physicians to find answers within complex medical texts. But that's not all! Google is now pushing the boundaries further by exploring multimodal capabilities. Imagine Med-PaLM 2 synthesizing patient information from images like chest x-rays or mammograms, revolutionizing diagnostic accuracy.The significance of this advancement cannot be understated. With the global healthcare data annotation tools market projected to reach a staggering USD 916.8 million by 2030, expanding at a remarkable rate of 27.5%, the potential impact on the industry is immense.At Label Your Data, we are contributing to these transformative advancements. Through ongoing projects, we label 100,000 blood cell images per day, providing invaluable insights for biotechnology companies. Additionally, we have experience annotating 50,000 medical records using NER annotation.Together, let's strive for a future towards better healthcare! 💙💻🏥#LabelYourData #MedicalAI #DataAnnotation #HealthcareInnovationhttps://lnkd.in/eu7XGQp3

14

🎉 Exciting News! 📢 Our latest research paper has just been published in#BiomedicalSignalProcessingAndControl(Q2) inElsevier! 📝🔬https://lnkd.in/de99tHkr

3

It seems that Google has chosen its niche. Multimodality is becoming the new normal, enabling AI to process diverse data types, from structured data and imaging to free text, waveforms, and more. In addition, LLMs allow for modular, reusable, and robust AI models to be trained with fewer manually labeled examples, making them more accessible and cost-effective for hospitals. It's an amazing time to live in!#ai #data #google #llms

5

Med PaLM 2, Google AI's latest medical language model, is here to revolutionize healthcare. Med PaLM 2 is trained on a massive dataset of medical literature and clinical data, making it the most comprehensive and up-to-date medical language model in the world.Med PaLM 2 can be used for a variety of tasks, including:- Answering medical questions: Med PaLM 2 can provide comprehensive and informative answers to a wide range of medical questions, from simple to complex.- Generating medical reports: Med PaLM 2 can generate accurate and well-written medical reports, saving clinicians time and effort.- Assisting with research: Med PaLM 2 can help researchers to identify new insights from medical data and to develop new treatments and diagnostic tools.To learn more about Med PaLM 2 and its potential to transform healthcare, watch this video:#MedPaLM2 #AI #Healthcare #Innovation

2

Today during #GoogleIO, we announced that we are working on the next evolution of our medical LLM: Med-PaLM 2 with multimodality.Effectively practicing medicine requires ingesting multiple formats of information, including texts and images. Research on the multimodal version of Med-PaLM 2 will help to synthesize and communicate information from images like chest X-rays, mammograms, and more to help doctors provide better care.We believe AI systems that leverage these data at scale using self-supervised learning with careful consideration of privacy, safety, fairness, and ethics will be the foundation of the next generation of learning health systems that scale world-class healthcare to everyone.We are working hard to validate this research, and I hope to see a future where our work helps fellow healthcare professionals provide even better patient care.https://lnkd.in/gJGYpCRH #google

54

I haven't been posting much since deep-diving into everything LLM for healthcare. Ugh, the intensity of learning, the speed of acquired knowledge becoming obsolete... I haven't experienced anything like this ever before. Today, I came across a Practical Guide for Medical LLMs, which definitely deserves sharing. It's a comprehensive review of publications and datasets, a must-read for anyone wanting to use LLM in their medical applications. Great work Michael Zhou et al. Here is a link: https://lnkd.in/e-qWKshW

199

14 Comments

Google has announced the release of Med-PaLM 2, a new AI system that can answer medical questions accurately and safely. Med-PaLM 2 is a large language model (LLM) trained on a massive dataset of medical text and code. This training allows Med-PaLM 2 to understand and respond to a wide range of medical questions, that are complex or challenging.In a recent study, Med-PaLM 2 was shown to be able to answer medical questions with an accuracy of 85%. This is on par with the precision of expert doctors. Med-PaLM 2 was also shown to be able to provide accurate and safe medical advice.Med-PaLM 2 has the potential to revolutionize the way that healthcare is delivered. It can be used to provide doctors with more accurate and up-to-date information, to help patients understand their medical conditions, and to develop new treatments and therapies.Here are some of the benefits of using Med-PaLM 2:Increased accuracy: Med-PaLM 2 can answer medical questions with an accuracy of 85%, which is on par with the accuracy of expert doctors.Improved safety: Med-PaLM 2 is able to provide accurate and safe medical advice.Increased efficiency: Med-PaLM 2 can help doctors to save time by providing them with access to up-to-date information and by helping them to diagnose and treat patients more quickly.Improved patient care: Med-PaLM 2 can help doctors to provide better care for their patients by providing them with more accurate information and by helping them to develop more effective treatments.If you are a doctor or a healthcare professional, I encourage you to learn more about Med-PaLM 2. It is a powerful new tool that can help you to improve the quality of care that you provide to your patients.Reference: https://lnkd.in/d_jdxhpjhttps://lnkd.in/dAYSS4jvhttps://lnkd.in/dG4WPfqj #ai #healthcare #google #medical #healthcare

13

上个星期,我们"医疗保健和人工智能英雄"阅读小组中,我的同事 Payam Khorramshahi 介绍了 Hu Huafeng et al. (2023)。Ye Ruijie, Jeyan Thiyagalingam, Frans Coenen, Jionglong Su#HeroesInHealthcareAITriple-kernel gated attention-based multiple instance learning with contrastive learning for medical image analysishttps://lnkd.in/g_pBJZvY

13

Build cohorts of students who will learn and thrive together in your medical programs. Learn how #AI can help you admit students who will make the greatest contributions.

2